Create app.py

app.py

@@ -0,0 +1,23 @@
+import gradio as gr
+import tensorflow as tf
+from wav2vec2 import Wav2Vec2Processor, Wav2Vec2ForCTC
+
+
+if __name__ == '__main__':
+    processor = Wav2Vec2Processor(is_tokenizer=False)
+    tokenizer = Wav2Vec2Processor(is_tokenizer=True)
+    model = Wav2Vec2ForCTC.from_pretrained("vasudevgupta/gsoc-wav2vec2-960h")
+
+    def _forward(speech: tf.Tensor):
+        speech = processor(speech)[None]
+        tf_out = model(speech, training=False)
+        return tf.squeeze(tf.argmax(tf_out, axis=-1))
+
+    def recognize_text(inputs):
+        _, speech = inputs
+        speech = tf.constant(speech, dtype=tf.float32)
+        speech = tf.transpose(speech)
+        tf_out = _forward(speech)
+        return tokenizer.decode(tf_out.numpy().tolist())
+
+    gr.Interface(fn=recognize_text, inputs="audio", outputs="text").launch()